source: https://doi.org/10.7892/boris.5770 | downloaded: 5.10.2021 1 [email protected] Correspondence: * influencing to characters access ornamentation over or weapons [11,12], competition [8-10], females male-male in and spiders involved species-speci- of often genitalia the processes fic are speciation examples in Well-known involved other [1-7]. be than might rapidly and more traits evolve traits reproduction that to suggests linked evidence theoretical and Empirical Background Klopfstein Seraina ) , (, in parasitoid courtship diplazontine antennal of evolution The http://www.biomedcentral.com/1471-2148/10/218 Klopfstein Switzerland aua itr uem(netbae) ensrse1,C-05Bern, CH-3005 15, Bernastrasse (Invertebrates), Museum History Natural EERHATCEOe Access Open ARTICLE RESEARCH osrit aeifune orsi eaiua risi iia a nohrgroups. evolutionary other whether in and way needed timescales, similar are larger a studies on in Further irreversible traits of wasps. is behavioural loss parasitoid have coiling courtship the of constraints antennal influenced that group of evolutionary have suggests this loss that constraints This in the therefore taxa. courtship whether and antennal studied ascertain here, at of the to evolved examined evolution of timescale Diplazontinae the any the in influenced in on coiling greatly reacquired irreversible Antennal never is courtship. was coiling during and antennal antennae rate, the low of comparatively use parsimony, a the maximum and by modifications independent supported antennal two is on coiling lost antennal is regaining Conclusions: approaches. coiling of Bayesian Antennal low rate and evolution. zero implying likelihood of A phylogeny, maximum speed reacquired. molecular antennal both low never the of that comparatively and with possession shows a occasions, congruent the reconstruction and highly that state homoplasy each are show Ancestral of coiling; coiling we levels behaviour. of antennal approach, coiling absence and comparison antennal the modifications model with or antennal a correlated double-coiling Using highly single-coiling, genera. is either multiple modifications find in we present coiling, is antennal state for As the we antennae. study phylogeny, the to system. well-resolved modifications i.e. character a morphological trait, complex of associated Results: behavioural this basis and courtship of the coiling peculiar On evolution antennal a of wasps. character of on mode parasitoid of history data diplazontine irreversibility evolutionary retrieve of instance the to species for reconstruct method 56 constraints, On both proposed for severe isolation. in recently coiling, under reproductive adaptations a antennal be in behavioural use might role we and evolution central Here, morphological their a loss. involve that plays often suggests which traits this behaviour behavioural sexes; courtship courtship for hand, case other the the especially is This speed. Background: Abstract tal et u td eel ag aito nsae oainadutasrcueo aeseii oiiain on modifications male-specific of ultra-structure and location shape, in variation large a reveals study Our . M vltoayBiology Evolutionary BMC u td rvdstefrtcmaaieeiec o ih orlto ewe male-specific between correlation tight a for evidence comparative first the provides study Our speitdb hoy risascae ihrpouto fe vlea oprtvl high comparatively a at evolve often reproduction with associated traits theory, by predicted As 1,2* oadL Quicke LJ Donald , 00Kofti ta;lcne iMdCnrlLd hsi nOe cesatcedsrbtdudrtetrso h Creative the of terms the cited. under properly distributed is article and work Access distribution, original Open use, the an unrestricted provided is permits This medium, which Ltd. any in Central (http://creativecommons.org/licenses/by/2.0), BioMed reproduction License licensee Attribution al; et Commons Klopfstein 2010 © 2010, 10 3,4 :218 hita Kropf Christian , 1 ecurd hsi oeie eerdt sDollo be as cannot to lost referred once sometimes traits is some This that reacquired. irreversi- fact is the evolution or constrained bility, of The mode [17,21-23]. obvious attention most less gained has behaviour constraints studie courtship extensively shaping been in has diversifying selection the sexual While of [16-20]. role place morphologi- take before adaptations long be cal selection might sexual which under behaviour, modified courtship morphologi- is dimorphic characters evolve sexually cal might than that quickly complex more trait even A [13-15]. choice female ,terl fevolutionary of role the d, ’ law, s Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 2 of 12 http://www.biomedcentral.com/1471-2148/10/218

and the characters involved as Dollo characters [21,24]. phylogeny of the group [[47], unpublished data], we Typically, Dollo characters are either very complex or assess the correlation between tyloids and antennal involve modifications at different levels, e.g. morphology coiling in a comparative framework. We reconstruct the and behaviour. Courtship related traits could represent evolutionary history of tyloid morphology, tyloid loca- such character systems, often involving specialized mor- tion and antennal coiling under maximum parsimony, phological adaptations and different behavioural compo- maximum likelihood and Bayesian approaches. Finally, nents from both the male and female sex. Under the we investigate whether the tyloid/antennal coiling constraint argumentation, courtship characters, as long character system shows signs of accelerated or of as they invoke also morphological adaptation, might constrained evolution on the phylogeny. thus evolve rather slowly, and moreover follow restricted pathways, with some characters only being lost and Results never reacquired during the evolutionary history of a Variation in tyloid morphology and coiling behaviour group. Of the 64 diplazontine species studied here, 34 have Courtship behaviour has been studied extensively in a male-specific antennal modifications (Table 1). These wide range of taxa, but with a strong bias towards verte- tyloids show a large variability in terms of their shape, brates. To understand the role of sexual selection in the coloration and location on the antenna (Fig. 1, Table 2), diversification of life, it is important to study especially with the shape ranging from narrowly linear tyloids, diverse groups, such as the order Hymenoptera. stretching over the whole length of the antennal seg- In this order, many species from a wide range of ment, to broadly oval or drop-shaped, short tyloids families have peculiar, male-specific modifications of the (Table 1). Most of the tyloids were plain coloured, antennae [25-32]. Many such modifications have been usually sharing the colour of the surrounding area, but shown to be connected to internal glands secreting a some species possess two-coloured tyloids with a yellow contact pheromone, and in some species, there is and a dark brown half. In other ichneumonid and bra- experimental evidence for antennal courtship mediating conid subfamilies, only a single or at most two different mate acceptance [26,33]. However, descriptions of types of tyloids can be found [27]. SEM investigations hymenopteran courtship behaviour are rather scarce. In revealed numerous pores on the surface of those tyloids most species, courtship can only be observed if virgin in 7 of 8 examined species with tyloids. In Enizemum females are at hand, which often requires rearing. In ornatum, a species with broadly linear tyloids specific to parasitoids, which account for the vast majority of the , we could not detect any pores (Figs. 1B and hymenopteran species, this may also necessitate rearing 1D); this brings into question whether the tyloids in this the host as well. This is either very intricate or even species are also used to transfer a secretion during impossible for species with unknown host relationships. courtship, or whether they rather serve for mechanical Courtship data are consequently very sporadic and stimulation. The number and positions of tyloids on the biased towards those species that can easily be cultured antennae are correlated, with a low number of tyloids [29,34-44]. However, to conduct comparative studies always being located on the middle segments of the and to reconstruct the evolution of courtship behaviour, antennae, and larger numbers of tyloids either ranging an extensive taxon sampling is needed. from the basal segments to the middle or from the mid- Recently, Steiner et al. [45] proposed a simple method dle close to the apex (Table 1). to reproduce a specific form of courtship-related anten- We performed the coiling test with 22 female speci- nation, i.e. antennal coiling, in museum specimens. This mens from 11 species and 110 male specimens from 56 method relies on the fact that the movement of the species. There was a clear sexual dimorphism in the antennal flagellum is achieved by differences in haemo- reaction to the coiling test. While the antennae of all lymph pressure and in the elasticity of the intersegmen- females bent slightly at the middle, the configurational tal membranes, since muscles are missing from changes observed in the male specimens varied a lot flagelliform antennae of all insects. This coiling test acts between genera and species. They ranged from an even as a proxy for courtship behaviour, as demonstrated in curve of the antenna (no coiling, Fig. 2A), to a single, four hymenopteran species with known courtship beha- tight coil in the middle of the antenna (single-coiling, viour [26,45,46]. Here we use this test to obtain anten- Fig. 2B) and two consecutive turns (double-coiling, nal coiling data for 56 species of Diplazontinae, and Fig. 2C). Each coiling type was observed in at least two examine the antennal morphology of a total of 64 spe- genera, and no intra-generic variation was observed cies. Compared to other ichneumonid subfamilies [27], except in the genus Homotropus (Table 1). The convex the Diplazontinae exhibit a large variation in the mor- tyloids in all cases are located in a way that they face phology and location of male-specific antennal struc- towards the inside of the coils, where they would be in tures, the so-called “tyloids”.Usingawellsupported close contact with the female antennae. Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 3 of 12 http://www.biomedcentral.com/1471-2148/10/218

Table 1 Taxon sampling, tyloid morphology and coiling behaviour Species SEM N(m) N(f) # Tyloids Tyloid type Tyloid Coiling location Bioblapsis cultiformis 1 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Campocraspedon annulitarsis 1 0 absent - no coiling Campocraspedon caudatus SEM 1 0 absent - no coiling Diplazon annulatus 1 0 absent - no coiling Diplazon deletus 1 0 absent - no coiling Diplazon hyperboreus 2 0 absent - no coiling Diplazon laetatorius 0 0 absent - ? Diplazon neoalpinus 0 0 absent - ? Diplazon orientalis 1 0 absent - no coiling Diplazon pallicoxa 0 0* absent* - ? Diplazon pectoratorius SEM 2 1 0 absent - no coiling Diplazon scutatorius 2 0 absent - no coiling Diplazon tetragonus 1 0 absent - no coiling Diplazon tibiatorius 1 0 absent - no coiling Diplazon varicoxa 1 0 absent - no coiling Diplazon zetteli 2 0 absent - no coiling Diplazon sp. D 2 0 absent - no coiling Enizemum ornatum SEM 1 8 linear, broad, whole length of antennomere, apical double-coiling plain-coloured Fossatyloides gracilentus SEM 1 4 linear, narrow, whole length of antennomere, middle single-coiling plain-coloured, with adjacent hole Homotropus crassicornis 0 11 linear, narrow, whole length of antennomere, basal ? plain-coloured Homotropus crassicrus 1 1 9 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus elegans 1 9 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus longiventris 1 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus melanogaster 2 1 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus nigritarsus 1 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus nigrolineatus 2 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus pallipes 1 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus pictus SEM 1 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus signatus 2 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus cf. lissosoma 1 6 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Homotropus subopacus 0 8 linear, narrow, whole length of antennomere, apical ? plain-coloured Homotropus vitreus 1 4 linear, narrow, whole length of antennomere, middle single-coiling plain-coloured Phthorima compressa 1 9 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Phthorima xanthaspis 2 8 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Promethes bridgmani 1 1 5 linear, narrow, whole length of antennomere, middle single-coiling plain-coloured Promethes melanaspis 1 4 linear, narrow, whole length of antennomere, middle single-coiling plain-coloured Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 4 of 12 http://www.biomedcentral.com/1471-2148/10/218

Table 1 Taxon sampling, tyloid morphology and coiling behaviour (Continued) Promethes nigriventris 0 4 linear, narrow, whole length of antennomere, middle ? plain-coloured Promethes sulcator 1 3 linear, narrow, whole length of antennomere, middle single-coiling plain-coloured Sussaba aciculata SEM 1 4 drop-shaped, shorter than antennomere, middle single-coiling plain-coloured Sussaba tertia 1 4 oval, broad, shorter than antennomer, middle single-coiling two-coloured Sussaba cognata SEM 10 2 0 absent - single-coiling Sussaba dorsalis 1 5 oval, broad, shorter than antennomer, middle single-coiling two-coloured Sussaba erigator SEM 20 1 5 drop-shaped, shorter than antennomere, middle single-coiling plain-coloured Sussaba flavipes SEM 6 3 4 drop-shaped, shorter than antennomere, middle single-coiling plain-coloured Sussaba placita 1 3 linear, narrow, shorter than antennomere, middle single-coiling plain-coloured Sussaba pulchella SEM 10 5 5 oval, broad, shorter than antennomer, middle single-coiling two-coloured Sussaba punctiventris 0 5 6 linear, narrow, shorter than antennomere, middle ? plain-coloured Sussaba roberti 1 4 drop-shaped, shorter than antennomere, middle single-coiling plain-coloured Syrphoctonus desvignesii 1 11 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Syrphoctonus fissorius 0 10 linear, narrow, whole length of antennomere, basal ? plain-coloured Syrphoctonus tarsatorius SEM 4 1 9 linear, narrow, whole length of antennomere, apical double-coiling plain-coloured Syrphophilus asperatus 1 0 absent - no coiling Syrphophilus bizonarius 1 0 absent - no coiling Syrphophilus tricinctorius SEM 1 0 absent - no coiling Syrphophilus tricinctus 1 0 absent - no coiling Tymmophorus erythrozonus 1 0 absent - no coiling Tymmophorus obscuripes 1 0 absent - no coiling Tymmophorus suspiciosus 1 0 absent - no coiling Woldstedtius biguttatus 1 0 absent - no coiling Woldstedtius citropectoralis 1 1 0 absent - no coiling Woldstedtius flavolineatus 1 0 absent - no coiling Woldstedtius holarcticus 1 0 absent - no coiling Woldstedtius sp. A 2 0 absent - no coiling Xestopelta gracilima 1 0 absent - no coiling Abbreviations: N(m)/N(f): number of individuals of the two sexes subjected to the antennal coiling test. # Tyloids: number of antennal segments carrying tyloids. Tyloid type: tyloid type described from shape, length and colour. Tyloid location: location of the tyloid-bearing antennal segments on the antennae; basal = segments 4 to about10, middle = segments 8 to 13, apical = segments 9 to 17. Coiling: coiling type as described in the materials and methods section. *information obtained from Manukyan 1987.

Ancestral state reconstructions The location of those tyloids cannot be unambiguously Ancestral state reconstructions of tyloid morphology by reconstructed, with a middle position being favoured by maximum likelihood and Bayesian MCMC methods are maximum likelihood and an apical position favoured by similar, but with some differences especially in the the Bayesian analysis. reconstruction of the tyloid shape (Table 2). Both meth- The MP, ML and Bayesian reconstructions of ances- ods reveal that the most recent common ancestor of the tral states in antennal coiling are shown in Figure 3. studied diplazontines probably had narrowly linear, long The three genus groups recognized by Klopfstein et al. and plain-coloured tyloids such as in Promethes, Syr- [47] exhibit different types of antennal coiling. The phoctonus, Homotropus and related genera (Table 2). members of the most basal Sussaba genus group all Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 5 of 12 http://www.biomedcentral.com/1471-2148/10/218

A B

C D

E F

G H

Figure 1 Variation in tyloid morphology in Diplazontinae. Shape of tyloids found in four species of diplazontine wasps. (A) and (C) Syrphoctonus tarsatorius, antennal segments 9 to 15 (A) and 10 (C), showing narrow, linear tyloids with abundant pores. Scale bars = 200 μm and 30 μm. (B) and (D) Enizemum ornatum, antennal segments 8 to 14 (B) and 10 (D), showing broad, linear tyloids with no pores. Scale bars = 200 μm and 50 μm. (E) and (G) Sussaba erigator, antennal segments 6 to 12 (E) and 8 (G), showing drop-shaped tyloids with pores. Scale bars = 200 μm and 50 μm. (F) and (G) Sussaba pulchella, antennal segments 6 to 14 (F) and 10 (G), showing oval, two-coloured tyloids with pores. Scale bars = 200 μm and 50 μm.

perform single-coiling. The ancestor of the Syrphoctonus which converged on the dependent model with 99.98% clade probably performed double-coiling, with coiling posterior probability (Bayes factor = 9.8). The correla- subsequently lost in the genus Woldstedtius and trans- tion between tyloid possession and antennal coiling was formed to single-coiling on two separate occasions, in almost strict, with Sussaba cognata as the single excep- Fossatyloides gracilentus and Homotropus vitreus.The tion. This species lacks tyloids or similar structures (as Diplazon genus group probably lost any form of anten- confirmed by SEM, Table 1), but consistently displays nal coiling during courtship already in its stem lineage. single coiling when examined with the coiling test. Overall, the rate of evolution of both morphological The evolution of tyloids and antennal coiling and behavioural characters was rather low, as indicated To test whether tyloids and antennal coiling evolved in in Figure 3 for antennal coiling. The consistency (CI) a correlated manner, we compared the likelihoods of a and retention index (RI) of presence and absence of model assuming independent evolution to a model both tyloids and coiling behaviour indicated low levels assuming that the two traits co-evolve [48]. The likeli- of homoplasy (CI = 0.333 and RI = 0.931 for tyloids and hood ratio test revealed significant support for the CI = 0.500 and RI = 0.960 for coiling behaviour). To dependent model (LHR = 15.43, df = 4, p = 0.0039). further study the mode of evolution of antennal coiling This result was confirmed by the Bayesian approach in Diplazontinae, we examined more closely the Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 6 of 12 http://www.biomedcentral.com/1471-2148/10/218

Table 2 Ancestral state reconstruction of tyloid morphology and location on the antenna in the ancestors of Diplazontinae and of the three genus groups Trait Diplazontinae Sussaba group Syrphoctonus group Diplazon group absence/presence absent 0.00/0.01 0.00/0.00 0.00/0.00 0.99/1.0 present 1.00/0.99 1.00/1.00 1.00/1.00 0.0/0.0 shape absent 0.05/0.02 0.03/0.01 0.02/0.00 0.99/0.90 narrow-linear 0.91/0.40 0.91/0.31 0.98/0.60 0.01/0.00 broad-linear 0.01/0.17 0.00/0.14 0.00/0.21 0.0/0.04 drop-shaped 0.01/0.16 0.01/0.18 0.00/0.12 0.0/0.04 oval 0.03/0.26 0.06/0.36 0.00/0.07 0.0/0.02 length absent 0.09/0.01 0.05/0.01 0.01/0.00 0.99/0.99 short 0.09/0.08 0.17/0.32 0.00/0.00 0.00/0.00 long 0.83/0.91 0.79/0.67 0.99/1.00 0.01/0.00 colour plain 1.00/1.00 1.00/1.00 1.00/1.00 1.00/1.00 two-coloured 0.00/0.00 0.00/0.00 0.00/0.00 0.00/0.00 location absent 0.19/0.02 0.03/0.01 0.07/0.00 0.99/0.99 middle 0.43/0.27 0.90/0.93 0.02/0.00 0.00/0.00 basal 0.34/0.63 0.06/0.03 0.89/0.95 0.01/0.00 apical 0.04/0.08 0.01/0.03 0.01/0.04 0.00/0.01 The probability of observing the indicated trait at the ancestral node of the specified groups is given first as resulting from the maximum likelihood and second from the Bayesian analysis. estimated transition rates between the two states, espe- and the movement actually performed by the males in cially, whether the rate of change from the state ‘absent’ Syrphoctonus tarsatorius. Three other hymenopterans for to ‘present’ (rate0- > 1) was equal to zero. To this end, which information on antennal courtship is available we restricted the transition rates to being either all further confirm this link [[45], unpublished results]. equal or one of them being zero, and compared the out- However, we have to bear in mind that species not show- comes to the unrestricted case with two different transi- ing any coiling in the test might still perform another tion rates. Under the unrestricted model, the rate0- > 1 form of antennal courtship, such as antennal tapping or was estimated to be zero, and this model had the high- stroking. Moreover, the link between the outcome of the est likelihood. However, when compared to the equal test and the behaviour actually displayed during court- rates model, this increase in likelihood was not signifi- ship might not be strict, as the coiling test only reflects cant with the likelihood ratio test (LRH = 1.61, df = 1, the presence of the morphological requirements to per- p = 0.20). The posterior probability of rate0- > 1 being form antennal coiling. In Sussaba cognata,theonlyspe- zero was estimated as 72% by the MCMC approach, cies in the Sussaba genus group that does not possess which integrates over phylogenetic uncertainty. When any tyloids but still performs antennal coiling in the coil- taking the mode of antennal coiling into account, ML ing test, one could imagine that the morphological estimationsconfirmazerorateforthetransitions0-> requirements for coiling have not yet been completely 1 and 0- > 2, as shown in Figure 4A. The corresponding reduced, although the tyloids as release and spread struc- posterior probability distributions obtained from the tures for a courtship pheromone are absent. It would be Bayesian analysis are shown in Figure 4B. of great interest to know whether the males of this spe- cies actually do or do not perform antennal coiling dur- Discussion ing courtship. Such data would provide insights into how We successfully applied the coiling test developed by the different aspects of the tyloid/antennal coiling system Steiner et al. [45] to obtain courtship behavioural data for co-evolve, i.e. the behaviour of the males per se, the mor- a large number of parasitoid species to reconstruct phological prerequisites for the movement, the tyloids as the evolution of this trait. The underlying assumption release and spread structures for a sex pheromone and was that the coiling test performed on dead specimens finally the preference of the females. The observation provides information about the mode of antennal coiling that not all aspects of this character complex need to be actually performed by the wasps. The observed sexual present in all species opens up the possibility of evolution dimorphism in the outcome of the antennal coiling test through intermediate stages. Enizemum ornatum pro- supports this assumption, as does the high correspon- vides another such example. Although this species pos- dence between the configurational change of the antenna sesses tyloids, we could see no pores on their surface Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 7 of 12 http://www.biomedcentral.com/1471-2148/10/218

Figure 2 Antennal coiling test applied to three diplazontine species. The change in configuration of the antennae is shown in three species when transferred from absolute ethanol (left column) to distilled water (right column). (A) Syrphophilus bizonarius. (B) Sussaba pulchella. (C) Syrphoctonus longiventris.

unlike in all the other species examined. The absence of establish the link between morphology and behaviour in tyloid pores might indicate that they no longer serve as the case of antennal courtship [45], and shed light on release and spread structures for a courtship pheromone. the evolution of courtship behaviour in this order. Instead, they could be involved in purely mechanical sti- The overall low rate of evolution and low levels of mulation in this species. homoplasy in the courtship traits analyzed here seem to The strong correlation observed in diplazontines contradict previous assertions that characters under sex- between tyloids and antennal coiling further supports ual selection would evolve at comparatively high rates the hypothesis that male-specific antennal modifications [1-4,6,9]. However, there are numerous previous exam- are likely involved in courtship [25,26,29,31-33,39,42,49]. ples demonstrating that parts of courtship behaviour Various forms of antennal courtship have been can evolve in a very conservative mode, translating in described in this diverse insect order, and even more high phylogenetic informativeness of some courtship diverse antennal modifications. As exemplified in the characters [50-53]. Although sexual selection can in present study, the coiling test could be used to further some cases lead to accelerated evolution, this does not Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 8 of 12 http://www.biomedcentral.com/1471-2148/10/218

Figure 3 Ancestral state reconstruction of antennal coiling in Diplazontinae. Ancestral state reconstructions of antennal coiling is shown on top of the phylogeny. Parsimony reconstruction is shown as branch colour, with multi-coloured branches representing equally parsimonious solutions. Ancestral state reconstructions are shown as pie diagrams on each node of interest, with the outer circle representing maximum likelihood and the inner circles the Bayesian reconstruction. The character states found in the terminal taxa are indicated as squares of the respective colours, with question marks indicating species for which we could not obtain data on coiling behaviour. Symbols after taxon names indicate presence (+) or absence (-) of tyloids. Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 9 of 12 http://www.biomedcentral.com/1471-2148/10/218

involved 31 species in two groups, implying that there wouldhavebeenplentyoftimeforare-acquisitionof the coiling state. On the other hand, there is a large var- iation found in antennal modifications in this subfamily, especially in the Sussaba genus group, implying that a more detailed examination of courtship behaviour in Diplazontinae might reveal a larger variation and higher evolutionary lability in other courtship traits. The concept of irreversibility of evolution was already introduced by Dollo in 1893 [24], and has been exten- sively disputed in the past [e.g. [55,56]]. It has been invoked to explain patterns of evolution of complex mor- phological characters, of sex determination systems, poly- ploidy and parthenogenesis, of gains and losses of genes, and many others [56-59]. In our case, there are argu- ments both in favour and against antennal coiling being a Dollo character in the narrow sense. From a theoretical point of view, a complex trait involving different charac- ter systems is difficult to be regained, once it is lost, and this description clearly applies to antennal coiling. On the other hand, there might be some pre-adaptations that could favour the reacquisition of some form of antennal courtship in parasitoid hymenopterans. For instance, the high sensitivity of the female antenna which is needed for host location and evaluation could facilitate the evolution of male behaviours that target the antenna of the female for sexual stimulation. Moreover, in the mounted position, the male body parts that can most easily access the female’s antennae are the male’s antennae. The varia- bility of antennal courtship and related morphological adaptations in the order Hymenoptera further suggest that these behaviours might not all be homologous, but instead might represent convergent evolutionary events. Further studies on a larger evolutionary scale are needed Figure 4 Estimated transition rates between coiling states.(A) to ascertain the irreversibility of a loss of antennal Transition rates between the three states of coiling behaviour as courtship in Hymenoptera. estimated by maximum likelihood. (B) Posterior distribution of rates obtained from Bayesian analysis. States are: 0 = no coiling, 1 = single coiling, 2 = double coiling. Conclusions Antennal coiling is a mode of courtship found in many hymenopteran and even other insect species, and repre- need to be the case, and might largely depend on the sents a complex character involving both morphological variability of a trait and on geographic and population and behavioural adaptations. We here demonstrate in a dynamic patterns [16]. Moreover, it has been shown comparative context that antennal courtship is highly that female preference sometimes even imposes selec- correlated with the possession of sexually dimorphic tion against a change in mating behaviour, thus deceler- antennal modification in diplazontine parasitoid wasps. ating the speed of evolution [e.g. [54]]. Courtship Moreover, we show that antennal coiling evolves at a behavioural traits are accordingly influenced by a variety low rate. It is lost two times independently in the sub- of selective pressures and evolutionary constraint which family and never re-acquired, which is in accordance might differ between traits in a single species. In our with Dollo’s law of irreversibility of evolution. case, the estimation of a zero rate of regain of antennal coiling suggests that this trait is under severe constraints Methods and that its loss evolves as an irreversible character, at Tyloid morphology least over the timescale examined. This is further sup- Presence and absence of tyloids and morphological types ported by the fact that the loss of antennal coiling were recorded for 64 species of Diplazontinae, Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 10 of 12 http://www.biomedcentral.com/1471-2148/10/218

classifying all male-specific, convex antennal structures four molecular markers, two mitochondrial (cytochrome of a minimum size of one fifth of the diameter of an oxidase subunit 1 and NADH 1) and two nuclear (28S, antennal segment as tyloids, following the conception of elongation factor 1-a). The phylogeny is well resolved, Diller [60]. To study the evolution of the tyloids, we with most nodes highly supported. We used the tree classified them into different types, examining their with the highest likelihood as obtained from the parti- number, length, shape and colour. To indicate the loca- tioned Bayesian analysis for the maximum parsimony tion of the tyloids on the antennae, we progressively (MP) and maximum likelihood (ML) estimations, and numbered the antennal segments from the segment 10’000 trees from the post-burnin Bayesian tree distribu- attached to the head capsule (scape) to the apex follow- tion for the MCMC methods. MP and ML reconstruc- ing Bin and Vinson [31]. The use of a stereo-microscope tion of ancestral states were conducted in Mesquite (Leica Wild M10) at magnification 80 proved sufficient [61], ML under the Mk1 model of evolution. Addition- to score these data; however, to make sure not to have ally, rate parameters and ancestral states at each node of overlooked tyloid-like structures and to study the ultra- interest were reconstructed under a Bayesian approach structure, we further examined some species by scan- using the program BayesMultiStates from the Bayes- ning electron microscopy (SEM). To this end, the Traits package [62]. In the Markov-chain Monte Carlo antennae of males stored in 80% ethanol were mounted, approaches, we applied an exponential reversible-jump air dried and gold-sputtered. SEM studies were per- hyperprior within the interval of zero to 30 and set the formed with a Philips XL30 FEG scanning electron ratedev parameter, which controls the proposal of new microscope. No overlooked, tyloid-like structures were rate values, to 8. This resulted in an acceptance rate found by SEM analysis. between 20% and 29% in all analyses, which falls inside the recommended range. To draw inferences about the Reproducing antennal coiling modes of evolution, we restricted the rate parameters to Courtship behaviour has to date only been described in obtain the likelihoods of different models [63]. To test two diplazontine species [45,46]. To simulate a change of whether antennal coiling evolved as an irreversible char- pressure inside the antennae and thus reproduce antennal acter, we forced the rate parameters of the reversals to courtship in museum specimens, we used a method devel- zero, and restricted the most ancestral node to the ‘pre- oped by Steiner et al. [45]. We first cut off the antennae of sent’ state, following Goldberg and Igic [64]. The result- dried specimens or of specimens kept in 80% ethanol. The ing likelihood values were then compared under the amputated antennae were then put into 10% aqueous likelihood ratio test [63]. Alternatively, the posterior dis- potassium hydroxide (KOH) for 10 min at room tempera- tribution of the models was assessed using reversible- ture to macerate the inter-segmentary membranes. Anten- jump MCMC [48]. To test for correlated evolution of nae from dried specimens were then put into 80% ethanol tyloids and coiling behaviour in diplazontine wasps, we overnight. Afterwards, we placed them in 100% ethanol used BayesDiscrete from the BayesTraits package [48] for 10 min and finally transferred them to distilled water. under an independent and a dependent model of evolu- Because of different viscosities of the two liquids, this tion. Likelihoods obtained under the two models with transfer resulted in an overpressure inside the antenna 50 ML attempts per tree were compared by a likelihood that led to an elongation of the inter-segmentary mem- ratio test. Posterior probabilities of the dependent and branes and eventually to a configurational change. We independent models, and harmonic means of the likeli- recorded this change in the configuration of the antennal hoods for Bayes factor comparisons, were obtained by segments of male specimens of 56 diplazontine species reverse-jump MCMC, using the above-mentioned para- (Table 1). To test reproducibility, we included multiple meter settings. specimens for some of the species for which enough fresh material was available. Tests were regarded as successful if Acknowledgements the antennae performed a distinct movement after being We would like to thank Werner Graber for providing the SEM micrographs transferred to water, following the recommendations of and Salome Steiner for the providing details about courtship behaviour of Steiner et al. [45]. We classified those movements as Syrphotonus tarsatorius. We acknowledge Stefan Bachofner for technical support, and Mark Shaw (Edinburgh), Scott Shaw (National Science antennal coiling that involved a specific range of segments Foundation grant BSI-07-17458), Rejio Jussila (Turku), and the Swedish coiled up in a spiral-like form. We denoted spirals of 200° Malaise Trap Project for providing specimens for this study. A previous to 360° as single-coils and spirals of more than 360° as version of the manuscript was considerably improved by the comments of two anonymous reviewers. This study was partly supported by a grant from double-coils. the Roche Research Foundation to SK and by NE/C519583 grant from the Natural Environment Research Council to DLQ (and Prof. Andy Purvis). Reconstructing the evolution or antennal courtship Author details The test phylogeny employed in this study was obtained 1Natural History Museum (Invertebrates), Bernastrasse 15, CH-3005 Bern, from Klopfstein et al. (unpublished data). It is based on Switzerland. 2Division of Community Ecology, Institute of Ecology and Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 11 of 12 http://www.biomedcentral.com/1471-2148/10/218

Evolution, University of Bern, Baltzerstr. 7, CH-3012 Bern, Switzerland. 22. Schiestl FP, Cozzolino S: Evolution of sexual mimicry in the orchid 3Division of Biology, Imperial College London, Silwood Park Campus, Ascot, subtribe Orchidinae: the role of preadaptations in the attraction of male Berkshire SL5 7PY, UK. 4Department of Entomology, Natural History Museum, bees as pollinators. BMC Evol Biol 2008, 8:27. London SW7 5BD, UK. 23. Ferrer MM, Good-Avila SV: Macrophylogenetic analyses of the gain and loss of self-incompatibility in the Asteraceae. New Phytol 2007, Authors’ contributions 173:401-414. SK identified the Diplazontinae species, documented the tyloid morphology, 24. Dollo L: Les lois de l’évolution. Bul Soc Belge Géol Pal Hydr 1893, 7:164-166. performed the coiling tests and did the comparative analyses. DLJQ 25. Isidoro N, Bin F, Romani R, Pujade-Villar J, Ros-Farre P: Diversity and contributed to the discussion of results and to the interpretation of the function of male antennal glands in Cynipoidea (Hymenoptera). Zool Scr phylogenetic inferences of character evolution. CK contributed to the 1999, 28(1-2):165-174. functional interpretation of the sexually dimorphic characters and to the 26. Romani R, Isidoro N, Riolo P, Bin F, Fortunato A, Turillazzi S, Beani L: A new conception of this research. All authors revised the manuscript drafts, read role for antennation in paper wasps (Hymenoptera, Vespidae): antennal and approved the final manuscript. courtship and sex dimorphic glands in antennomeres. Insect Soc 2005, 52(1):96-102. Received: 9 February 2010 Accepted: 20 July 2010 27. Gokhman VE, Krutov VV: On external structure of male antennae in the Published: 20 July 2010 subfamily Ichneumoninae (Hymenoptera, Ichneumonidae) and related groups. Zool Zh 1996, 75(8):1182-1194. References 28. Isidoro N, Bin F, Colazza S, Vinson SB: Morphology of antennal gustatory 1. Ritchie MG: Sexual selection and speciation. Annu Rev Ecol Evol Syst 2007, sensilla and glands in some parasitoid Hymenoptera with hypothesis on 38:79-102. their role in sex and host recognition. J Hym Res 1996, 5:206-239. 2. Panhuis TM, Butlin R, Zuk M, Tregenza T: Sexual selection and speciation. 29. Romani R, Rosi MC, Isidoro N, Bin F: The role of the antennae during Trends Ecol Evol 2001, 16(7):364-371. courtship behaviour in the parasitic wasp Trichopria drosophilae. JExp 3. Arnqvist G, Edvardsson M, Friberg U, Nilsson T: Sexual conflict Biol 2008, 211:2486-2491. promotes speciation in insects. Proc Natl Acad Sci USA 2000, 30. Isidoro N, Bin F: Male antennal gland of Amitus spiniferus (Brethes) 97(19):10460-10464. (Hymenoptera: Platygastridae), likely involved in courtship behaviour. Int 4. Gavrilets S: Rapid evolution of reproductive barriers driven by sexual J Insect Morphol 1995, 24:365-373. conflict. Nature 2000, 403:886-889. 31. Bin F, Vinson SB: Morphology of the antennal sex-gland in male Trissolcus 5. Lande R: Models of speciation by sexual selection on polygenic basalis (Woll.) (Hymenoptera: Pteromalidae). Int J Insect Morphol 1986, characters. Proc Natl Acad Sci USA 1981, 78(6):3721-3725. 15:129-138. 6. Gage MJG, Parker GA, Nylin S, Wiklund C: Sexual selection and speciation 32. Guerrieri E, Pedata P, Romani R, Isidoro N, Bin F: Functional anatomy of in mammals, butterflies and spiders. Proc R Soc Lond B Biol Sci 2002, male antennal glands in three species of Encyrtidae (Hymenoptera: 269(1507):2309-2316. Chalcidoidea). J Nat Hist 2001, 35:41-54. 7. McPeek MA, Gavrilets S: The evolution of female mating preferences: 33. Bin F, Wackers F, Romani R, Isidoro N: Tyloids in Pimpla turionellae (L.) are differentation from species with promiscous males can promote release structures of male antennal glands involved in courtship speciation. Evolution 2006, 60(10):1967-1980. behaviour (Hymenoptera: Ichneumonidae). Int J Insect Morphol 1999, 8. Huber BA, Sinclair BJ, Schmitt M: The evolution of asymmetric genitalia in 28(1-2):61-68. spiders and insects. Biol Rev 2007, 82:647-698. 34. Eggleton P: Patterns in male mating strategies of the Rhyssini: a 9. Arnqvist G: Comparative evidence for the evolution of genitalia by holophyletic group of parasitoid wasps (Hymenoptera: Ichneumonidae). sexual selection. Nature 1998, 393:784-786. Anim Behav 1991, 41:829-838. 10. Huber BA: Rapid evolution and species-specificity of genitalia: 35. Beukeboom LW, van den Assem J: Courtship and mating behaviour of fact or artifact? Org Div Evol 2003, 3:63-71. interspecific Nasonia hybrids (Hymenoptera, Pteromalidae): a 11. Emlen DJ, Marangelo J, Ball B, Cunningham CW: Diversity in the weapons grandfather effect. Behav Genet 2001, 31(2):167-177. of sexual selection: horn evolution in the beetle genus Onthophagus 36. Beukeboom LW, van den Assem J: Courtship displays of introgressed, (Coleoptera: Scarabaeidae). Evolution 2005, 59(5):1060-1084. interspecific hybrid Nasonia males: further inverstigations into the ‘ ’ 12. Emlen DJ: The evolution of weapons. Annu Rev Ecol Evol Syst 2008, grandfather effect . Behaviour 2002, 139:1029-1042. 39:387-413. 37. Ovruski SM, Aluja M: Mating behavior of Aganaspis pelleranoi (Brèthes) 13. Seehausen O, Mayhew PJ, van Alphen JJM: Evolution of colour patterns in (Hymenoptera: Figitidae, Eucoilinae), a fruit fly (Diptera: Tephritidae) East African cichlid fish. J Evolution Biol 1999, 12:514-534. larval parasitoid. J Insect Behav 2002, 15(1):139-151. 14. Roulin A, Bize P: Sexual selection in genetic colour-polymorphic species: 38. Budrienné A, Budrys E: Behavioural elements influencing mating success a review of experimental studies and perspectives. J Ethol 2007, of Symmorphus allobrogus (Hymenoptera: Eumeninae). Acta Zool Litu 25:99-105. 2004, 14(2):39-47. 15. Cuervo JJ, Møller AP: Ecology and evolution of extravagant feather 39. Battaglia D, Isidoro N, Romani R, Bin F, Pennacchio F: Mating behaviour of ornaments. J Evolution Biol 1999, 12:986-998. Aphidius ervi (Hymenoptera: Braconidae): The role of antennae. Eur J 16. Podos J, Warren PS: The evolution of geographic variation in birdsong. Entomol 2002, 99:451-456. Advances in the Study of Behaviour Amsterdam: ElsevierBrockmann HJ, 40. Vinson SB: Courtship behavior and evidence for a sex pheromone in the Roper T, Naguib M, Wynne-Edwards K, Barnard C, Mitani J 2007, 37:403-458. parasitoid Campoletis sonorensis (Hymenoptera: Ichneumonidae). Environ 17. Podos J, Huber SK, Taft B: Bird song: The inferface of evolution and Entomol 1972, 1(4):409-414. mechanism. Annu Rev Ecol Evol Syst 2004, 35:55-87. 41. Ayasse M, Paxton J, Tengö J: Mating behavior and chemical 18. Spaulding A: Rapid courtship evolution in grouse (Tetraonidae): communication in the order Hymenoptera. Annu Rev Entomol 2001, contrasting patterns of acceleration between the Eurasian and North 64:31-78. American polygynous clades. Proc R Soc Lond B Biol Sci 2007, 42. Bin F, Strand MR, Vinson SB: Antennal structures and mating behavior in 274(1613):1079-1086. Trissolcus basalis (Woll.) (Hym.: Scelionidae), egg parasitoid of the green 19. Blomberg SP, Garland TJ, Ives AR: Testing for phylogenetic signal in stink bug. Proceedings of the 2nd International Symposium on Trichogramma comparative data: behavioral traits are more labile. Evolution 2003, and other egg parasites: 1988; Guangzhou (China) 1988, 143-151. 57(4):717-745. 43. Jachmann F, van den Assem J: The interaction of external and internal 20. Gittleman JL, Anderson CG, Kot M, Luh H-K: Phylogenetic lability an drates factors in the courtship of parasitic wasps (Hym, Pteromalidae). of evolution: a comparison of behavioral, morphological and life history Behaviour 1993, 125(1-2):1-19. traits. Phylogenies and the comparative method in animal behavior Oxford, 44. Cheng L, Howard RW, Campell JF, Charlton RE, Nechols JR, Ramaswamy SB: U.K.: Oxford University PressMartins EP 1996, 166-205. Mating behavior of Cephalonomia tarsalis (Ashmead) (Hymenoptera: 21. Bull JJ, Charnov EL: On irreversible evolution. Evolution 1985, Bethylidae) and the effect of female mating frequency on offspring 39(5):1149-1155. production. J Insect Behav 2004, 17(2):227-245. Klopfstein et al. BMC Evolutionary Biology 2010, 10:218 Page 12 of 12 http://www.biomedcentral.com/1471-2148/10/218

45. Steiner S, Kropf C, Graber W, Nentwig W, Klopfstein S: Antennal courtship and functional morphology of tyloids in the parasitoid wasp Syrphoctonus tarsatorius (Hymenoptera: Ichneumonidae: Diplazontinae). Arthropod Struct Dev 2010, 39(1):33-40. 46. Rotheray GE: Courtship, male swarms and a sex pheromone of Diplazon pectoratorius (Grav.). Entomol Gaz 1981, 32(1):193-196. 47. Klopfstein S, Kropf C, Quicke DLJ: An evaluation of phylogenetic informativeness profiles and the molecular phylogeny of Diplazontinae (Hymenoptera, Ichneumonidae). Syst Biol 2010, 59(2):226-241. 48. Pagel M, Meade A: Bayesian analysis of correlated evolution of discrete characters by reversible-jump Markov chain Monte Carlo. American Naturalist 2006, 167(6):808-825. 49. Romani R, Isidoro N, Riolo P, Bin F: Antennal glands in male bee: structures for sexual communication by pheromones? Apidologie 2003, 34:603-610. 50. Price JJ, Lanyon SM: Reconstructing the evolution of complex bird song in the Oropendolas. Evolution 2002, 56(7):1514-1529. 51. Puniamoorthy N, Feng-Yi Su K, Meier R: Bending for love: losses and gains of sexual dimorphisms are strictly correlated with changes in the mounting position of sepsid flies (Sepsidae: Diptera). BMC Evol Biol 2008, 8:155. 52. Tinbergen N: Comparative studies of the behaviour of gulls (Laridae): a progress report. Behaviour 1959, 15(1/2):1-70. 53. Kennedy M, Spencer HG, Gray RD: Hop, step and gape: do the social displays of the Pelecaniformes reflect phylogeny? Anim Behav 1996, 51:273-291. 54. Nowicki S, Searcy WA, Hughes M, Podos J: The evolution of bird song: male and female response to song innovation in swamp sparrows. Anim Behav 2001, 62:1189-1195. 55. Gould SJ: Dollo on Dollo’s law: irreversibility and the status of evolutionary laws. J Hist Biol 1970, 3:189-212. 56. Collin R, Miglietta MP: Reversing opinions on Dollo’s law. Trends Ecol Evol 2008, 23(11):602-609. 57. Collin R, Cipriani R: Dollo’s Law and the re-evolution of shell coiling. Proc R Soc Lond B Biol Sci 2003, 270:2551-2555. 58. Bridgham JT, Ortlund EA, Thornton JW: An epistatic ratchet constrains the direction of glucocorticoid receptor evolution. Nature 2009, 461:515-520. 59. Igic B, Bohs L, Kohn JR: Ancient polymorphism reveals unidirectional breeding system shifts. Proc Natl Acad Sci USA 2006, 103(5):1359-1363. 60. Diller EH: Klärung einiger Taxa der Gattung Sussaba Cameron, 1909 (Hymenoptera, Ichneumonide, Diplazontinae). Entomofauna 1980, 1(5):58-64. 61. Maddison WP, Maddison DR: Mesquite: a modular system for evolutionary analysis. Version 2.01. 2007 [http://mesquiteproject.org]. 62. Pagel M, Meade A, Barker D: Bayesian estimation of ancestral character states on phylogenies. Syst Biol 2004, 53(5):673-684. 63. Pagel M: Inferring evolutionary processes from phylogenies. Zool Scr 1997, 26(4):331-348. 64. Goldberg EE, Igic B: On phylogenetic tests of irreversible evolution. Evolution 2008, 62(11):2727-2741.

doi:10.1186/1471-2148-10-218 Cite this article as: Klopfstein et al.: The evolution of antennal courtship in diplazontine parasitoid wasps (Hymenoptera, Ichneumonidae, Diplazontinae). BMC Evolutionary Biology 2010 10:218.

Submit your next manuscript to BioMed Central and take full advantage of:

• Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit